On Tuesday 19 January 2021 15:50:49 Todd Zuercher wrote: > I under stand the analogy, but I wonder if Linuxcnc is erroneously > behaving as if the "leadscrew" were multiple start, with the number of > starts equal to the encoder's counts/rev. Resulting in the syncing of > the axis to the spindle is inconsistent if the syncing move isn't > always the same? (Not saying this is how it is, just proposing a > possible reason if the problem is real.) > I think in cases where the spindle encoder is quite low resolution, that it may well contribute. albeit probably not to that extent. Because on my g0704, the spindle encoder is actually a 1024 ppr on the rear of the motor, while the index is derived from an ATS-667 watching a screw glued to the side of the drawbar nut, which makes my spindle encoder have an effective count when the head is in low gear of something a bit north of 14,000. Just over 7000 in high gear. All fixed in hal.
I think a bigger reason for my thread timing wibbles in the G33.1 cycle is far more related to the sample rate imposed by the servo only thread. We need a 10 kilohertz thread just for reading the spindle encoder, for exactly this reason. Quantization noise IOW. Between that, and the fact that the post of this G0704 is a degree or more out of square with the base. Its leaning sideways just enough to detect it. Tramming the head square to the table, and you can see the error in the head to slider joint. Ad you can dial a square sitting on the table, and use up the range of a .0001" dial in less than 2" of vertical motion. I loosened the bolts a year back and pulled it sorta square, but then I can't tighten the bolts, they are seriously bound. So sometime when I've nothing better to do, I'll take it off, and run a reamer thru 3 of the holes the next size bigger, and then try to plumb it again when I put the bolts back in. Maybe this summer if I don't miss roll call first. But to do that really right, I'll have to invest in a cylindrical square. And to do it really right, I'll need some sort of a worm drive rigged for tramming the head. Face it, the G0704 is a Chinese piece of BBLB crap. I don't really understand why I keep screwing with it. My 6040 gantry is squarer that this piece of junk. And at the moment, its halfway thru making the first of 2 ER20 chuck wrenches out of 1/2" alu plate. Which will, sometime in the next day or so, generate some questions about G5.2, the nurbs thing. Figure I might as well make them purty. Its a bit funny, its been in LCNC since 2009, and its still labeled very experimental. Has anyone actually used it? > I don't do any threading or have any machines with a spindle encoder. > But I have observed this debate on whether this problem exists in > Linuxcnc for years now here on this list. This is true, for pushing 2 decades now. > Take care and stay well everybody. > Todd Zuercher > P. Graham Dunn Inc. > 630 Henry Street > Dalton, Ohio 44618 > Phone: (330)828-2105ext. 2031 Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis Genes Web page <http://geneslinuxbox.net:6309/gene> _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
